{
  "id": "2606.24551",
  "title": "GUI vs. CLI: Execution Bottlenecks in Screen-Only and Skill-Mediated Computer-Use Agents",
  "first_seen": "2026-06-26",
  "published_date": "2026-06-22",
  "observed_dates": [
    "2026-06-26"
  ],
  "score": {
    "novelty": 97,
    "practical_impact": 100,
    "technical_depth": 77,
    "implementation_potential": 99,
    "relevance": 100,
    "community_signal": 53,
    "summary_confidence": 85,
    "overall": 90,
    "weights": {
      "novelty": 0.2,
      "practical_impact": 0.2,
      "technical_depth": 0.15,
      "implementation_potential": 0.15,
      "relevance": 0.15,
      "community_signal": 0.1,
      "summary_confidence": 0.05
    }
  },
  "recommendation": "Read",
  "innovation_summary": "GUI vs CLI: Execution Bottlenecks in Screen-Only and Skill-Mediated Computer-Use Agents: We introduce a matched execution-layer benchmark of 440 desktop tasks across 18 applications and 12 workflow categories, where screen-only GUI agents and skill-mediated CLI agents receive.",
  "why_it_matters": [
    "Overall signal 90/100 driven by novelty 97 and practical impact 100.",
    "It maps to cross-cutting AI systems work even without explicit category metadata.",
    "Community signal includes 6 upvote(s) and 1 comment(s), which helps separate durable interest from title-only curiosity."
  ],
  "implementation_angle": [
    "Implementation potential scores 99/100; prioritize adaptation paths for internal agent, evaluation, or platform workflows.",
    "No linked repository is present, so expect more translation work before the ideas are production-ready.",
    "Technical depth scores 77/100, so a quick skim should focus on architecture, data, and evaluation sections before full adoption work."
  ],
  "caveat": "Evidence appears benchmark-centric, so verify transfer to production workloads before acting on the claims.",
  "links": {
    "hugging_face": "https://huggingface.co/papers/2606.24551",
    "arxiv": "https://arxiv.org/abs/2606.24551"
  }
}